Abstract During drug abstinence, re-exposure to cues previously associated with cocaine often trigger drug relapse. In rodent models of cocaine seeking and relapse, animals that self-administer cocaine in the presence of contingent cues often establish a strong association between cues and cocaine, such that after drug withdrawal, the presence of cues induces strong cocaine seeking. Cue-induced cocaine seeking is extremely long-lasting and intensifies progressively after withdrawal. The goal of this application is to develop new concepts and approaches through which the cue-cocaine association can be disrupted to reduce cocaine relapse. The cue-cocaine association that drives cocaine seeking shares general features of cue-conditioned memories. Similar to classic conditioning memories, cue-conditioned drug memories also undergo a destabilization and reconsolidation process after retrieval. During the brief destabilization time window, amnesic treatments are often more effective in reducing subsequent cue-induced cocaine seeking. However, the neural substrates that mediate cue-drug memory retrieval and reconsolidation remain elusive. Targeting this knowledge gap, this application focuses on cocaine-generated silent synapses and their dynamic changes within the basolateral amygdala (BLA) to nucleus accumbens (NAc) projection. We recently showed that cocaine self-administration generates silent synapses in the BLA-to-NAc projection. Silent synapses are excitatory synapses that contain NMDA receptors without stable AMPA receptors (AMPARs). Our additional results suggest that cocaine-generated silent synapses may serve as the initial hubs to establish a potentially new set of circuits. After cocaine withdrawal, BLA-to-NAc silent synapses become `un-silenced' by recruiting calcium-permeable AMPARs (CP-AMPARs), resulting in consolidation of the silent synapse-imbedded circuits. Reversing the un-silencing of BLA-to-NAc silent synapses decreases cue-induced cocaine seeking. These results suggest that the newly formed, silent synapse-embedded BLA-to-NAc projections contribute to the establishment and subsequent consolidation of cue-cocaine association. Based on extensive preliminary results, we hypothesize that after cocaine withdrawal, a brief re-exposure to cocaine-associated cues instantly induces CP-AMPAR internalization and re-silences the same set of BLA-to-NAc silent synapses that are generated by cocaine self-administration, contributing to the destabilization of cue-cocaine association. These re-silenced synapses are un-silenced again by re-recruiting CP-AMPARs hours after cue re-exposure, contributing to reconsolidation of cue-cocaine association. We will use electrophysiology, optogenetics, in vivo calcium imaging, viral-mediated gene transfer, and operant behavioral assays to test this hypothesis. By accomplishing the proposed work, we may identify a cellular and circuit basis underlying destabilization and reconsolidation of cue-cocaine association after retrieval and validate therapeutic angles to reduce cocaine relapse. Thus, objectives of this application are highly relevant to the missions of the NIDA, NIH.